Abstract
Herpes simplex virus-1 (HSV-1) is a ubiquitous pathogen that establishes lifelong latent infections in post-mitotic neurons, most commonly in the peripheral ganglia. Latent HSV-1 infection is asymptomatic, but the virus periodically reactivates, which can lead to significant disease, including a life-threatening encephalitis or recurrent HSV-1 ocular infection, a leading cause of infectious blindness worldwide. Recent evidence suggests that recurrent HSV-1 infection may also impact the progression of Alzheimer’s disease. Therefore, there is a need to understand how the virus remains latent in neurons to ultimately prevent reactivation and recurrent infection. During latency, the promoters of viral lytic genes are associated with heterochromatin, which is thought to maintain long-term gene silencing. However, HSV latency is heterogenous and latent viral genomes associate with different subnuclear structures and cellular proteins, which may result in different forms of latency that are more or less capable of reactivation. Additionally, this can be further impacted by the inflammatory environment and the presence of cytokines or other signaling molecules. In Chapter 2, we explore the role of type I interferons (IFNs) and promyelocytic leukemia nuclear bodies (PML-NBs) in promoting a restricted form of HSV latency. We first characterized PML distribution in primary peripheral neurons isolated from adult and postnatal mice and found that they are largely devoid of detectable PML-NBs. Treatment of these primary murine neurons with type I IFN induces robust formation of PML-NBs that continue to persist following cessation of IFN signaling. A large proportion of HSV-1 genomes are stably entrapped by PML-NBs throughout latency when IFNα is present during initial infection, and reactivation is restricted under these conditions. However, the ability of HSV to reactivate is rescued if PML is depleted either prior to or following infection, suggesting that IFNα-induced PML-NBs are required for the restriction of HSV-1 reaction. In Chapter 3, we further investigated the localization of latent viral genomes in primary murine neurons and found they colocalize not only to alpha-thalassemia/mental retardation syndrome x-linked protein (ATRX) in the context of PML-NBs, but also to regions of dense ATRX staining localized outside of PML-NBs, even in the absence of IFNs. Furthermore, depletion of ATRX in latently infected neurons in vivo decreased the latent viral load in trigeminal ganglia (TG) and increased lytic gene expression during reactivation of superior cervical ganglia (SCG). Interestingly, we show that the subnuclear positioning of a latent viral genome modulates not only its chromatin structure and association with histones, but also its compaction state, as superresolution microscopy revealed that PML-NB-associated genomes are significantly less compact than non-PML-NB-associated genomes. Lastly, we characterized PML and ATRX distribution in human neuronal cell models of HSV-1 latency and found that IFNα increases PML-NB number and colocalization of PML-NBs to latent HSV-1 genomes in these cells. Together, these studies have expanded our understanding of heterogeneity in latency and provide key insight into what form of latency is the most repressive (also known as deep latency), why it is more repressive and how different forms of latency arise so that we may be able to develop therapies that manipulate latency and drive it into its deepest form.
